Most proteins are evolutionarily optimized for function, but not for folding. Thus, understanding protein folding mechanisms will help us design proteins that are optimized for folding without altering their functions. In addition, misfolding and aggregation underlie fatal diseases such as cystic fibrosis, Alzheimer's disease and other amyloidoses, and type-II diabetes. A predictive framework for protein folding, particularly in the nonnative states and its sequence determination, will greatly impact biomedical research of folding-related diseases. We will develop and validate such a predictive framework for protein folding through the studies of kinetics and thermodynamics of a few fast-folding proteins. We intend to understand the determinants of folding mechanisms for these proteins at all-atom details and validate our understanding by comparing with experiment extensively. An all-atom molecular mechanics force field in both explicit solvent and in implicit solvent will be used to characterize the nonnative states and the folding pathways. [unreadable] [unreadable] 1) We plan to: (a) investigate the interplay of sequence and topology in the determination of the folding rates and pathways for two engineered proteins with Zn-finger motif: FSD1 and PDA8D; (b) perturb their folding pathways by mutation to achieve sub-microsecond folding rates in this family. 2) We will investigate: (a) what makes the predicted rate of protein A based on topology correct, but the predicted rate of Engrailed Homeodomain off by a factor of 40; (b) what makes protein A so sensitive to mutation -- a single mutation can change its rate by a factor of 10 or higher; (c) what makes Engrailed Homeodomain tolerate drastic changes in sequence with little change in folding rates. We will computationally probe many aspects of their folding processes, the denatured states, the transition states, the intermediate states if any, and the pathways towards native states to understand the differences. [unreadable] [unreadable]